[Webkit-unassigned] [Bug 146056] [GTK] [Wayland] Allow building and testing the Wayland target with the default JHBuild moduleset.

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=146056

--- Comment #9 from Carlos Garcia Campos <cgarcia at igalia.com> ---
(In reply to comment #7)
> (In reply to comment #3)
> > I think we should go straight to the latest stable release, 3.16.4, since
> > it's closer to what our users will actually have. There's less value in
> > testing with 3.12.x since it's already quite old.
> > 
> 
> I think that we may be risking to break compatibility with older versions
> without noticing.

However, what usually happens is the opposite, things stop working or are broken with recent versions of GTK, and we don't notice it, but our users do. Remember the huge scrollbars and other theming issues, the GMutexLocker build issue, etc.

> If we put 3.16.x on the JHBuild, we won't longer test with older versions.
> The possibility that a patch inadvertently introduces some change that
> breaks with < 3.16.x is real.

We just need to be careful to use conditional compilation when using newer GTK+ API. It's not that common. The thing is that what we are testing is far from real.

> But, the same will happen with an update to 3.12 (is only a matter of
> limiting the hypothetical breakage to <3.12 instead of <3.16).
> 
> In any case, I'm fine with updating GTK to 3.16 if you all are also ok.
> 
> 
> I'm wonder if along with the update to GTK-3.16 we should also (or we want
> to) update any other library that GTK may depend on like glib, cairo, etc?

Yes, indeed. We need to check if newer GTK+ also requires newer versions of other deps, I bet it requires newer glib, for example, but also ATK, pango, etc. The other important dependency that affects the layout tests output is cairo. So, I would take advantage that we will have to rebaseline to also update cairo. With recent enough cairo and GTK+ we have HiDPI support, that we have never tested because we were using an old GTK.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-unassigned/attachments/20150618/1f17ad2b/attachment-0001.html>